Machine learning in mechanics and materials

Co-Chairs:

Zhao Qin, Massachussets Institute of Technology

Markus J. Buehler, Massachussets Institute of Technology

 

Summary:

Advances in theory and computing enable us to explore innovative composition and synthesis strategies and develop unprecedented materials for a broad spectrum of functions. To determine whether a given structure is the best suited one, one cannot expect to solely use brute force to compare it with all the peer structures as what has been previously utilized conventionally, as the computational cost would be prohibitive for sophisticate designs. However, artificial intelligence (AI), enabled by machine learning (ML) technique, now provides a novel feasible way of solving this problem and offers an efficient approach to searching a high-dimensional parameter space for optimal solutions, driven by data generated from experiments or first-principles physics-based simulations. This symposium will emphasize scientific advances at this frontier, and explore how AI and ML can be used to help the design of new functional materials by optimizing the material functions through adjusting the compositions and multiscale structures that range from the atomistic to the macroscale.